Introduction

Biocompatibility of orthodontic appliances has been the subject of several studies, mainly due to the concern related to the possible discharge of potentially toxic and carcinogen ions during orthodontic treatment [1–3]. Assessment of trace elements in human might exhibit the risk associated with orthodontic appliances [2], which are made of alloys containing several metals [1, 2, 4], among which of the major concern are chromium and nickel [1–3, 5–7]. Both of these genotoxic, mutagenic, and cytotoxic metals might induce contact allergy, asthma, hypersensitivity, birth defects, and reproductive harms [1–3, 5–15]. Corrosion of orthodontic alloys might lead to release of sizeable amounts of nickel and chromium ions into saliva [5, 8, 10, 12]. Chromium is added to these alloys to form an anti-corrosive passive chromium oxide film [8, 16]. Nevertheless, in clinical situations, this protective layer is disrupted via several mechanisms, such as thermal stresses [3, 9, 17], saliva flow [10, 12, 17], mastication [18], brushing [12], biofilm layer [5, 9] and its by-products or enzymatic activities [3, 5, 11, 19–21], recycling of the appliances [9, 21], friction between brackets and wires [3], occlusal loadings [11, 20], as well as acidic drinks, mouthwashes, or toothpastes [5, 10, 12, 14, 17, 20, 21].

Considering the potential danger of these trace elements, it is essential to define concentrations of these ions released during orthodontic treatments [1, 2]. The majority of the studies in this regard have been conducted in vitro [1, 2, 5, 16] which is not a method relevant to clinical conditions since it cannot reproduce highly dynamic and complex oral environment [1–3, 5, 8–11, 13, 16]. Most of the few in vivo studies on this subject have been focused primarily on metal concentrations in saliva, which is the first diluent of the human body and also allows long-term analyses [10, 14, 16, 18, 22]. The reports have been quite controversial and thus confusing to the orthodontists [5, 9]. The dispute stems from several confounding variables such as genetics, patients’ dietary, and smoking habits in addition to differences in exposure to trace elements from other sources (e.g., air and water) [10, 18, 23, 24]. Moreover, improved sampling methods appear necessary for clarifying this topic. Most of in vivo studies have evaluated short-term exposures to orthodontic appliances (i.e., maximum 2 months) [9, 13, 22, 25, 26]—which is unrealistic because orthodontic treatments last for 2 or 3 years [1, 16]—and the few long-term studies were all retrospective [10, 14, 16, 18]. Among the long-term retrospective investigations, one was descriptive [14] and therefore not sufficiently reliable [5]. Of the remaining three long-term retrospective cohort studies, the treatment and control subjects had not been matched, apart from in merely one study [10]. Moreover, the results in long durations have been controversial [10, 16, 18].

The literature regarding salivary metal ion release from orthodontic appliances lacks any controlled prospective cohort studies for durations longer than 2 months. A long-term in vivo study seems to be essential for appropriately simulating orthodontic treatments besides reducing behavioral, environmental, and genetic confounders. Hence, we aimed to assess the concentrations of nickel and chromium in orthodontic patients immediately before treatment initiation (as the control) and 6 and 12 months after beginning of treatment.

Subjects and Methods

This prospective cohort study was performed on 120 specimens obtained from a cohort of 20 orthodontic patients. The exclusion criteria comprised the presence of any systemic diseases, any history of allergic reactions, medication intake, alcohol consumption or smoking, any caries [14], metal restorations such as amalgam fillings or fixed prostheses placed before or during the treatment, and having some teeth missing or extracted (excluding the third molars). The mean age of the included patients was 16.0 ± 4.6 years (range 14–23), and 12 participants were females. The ethics of the study protocol were approved by the internal review board of the university, and written consents were taken from the subjects or their parents after thorough oral and written explanation. The fixed appliances at the time of sample collection consisted of bonded 0.018 in. slot pre-adjusted Roth prescription stainless steel brackets (Discovery, Dentaurum, Pforzheim, Germany) on all teeth except the molars and an average of four to eight stainless steel orthodontic bands (Beijing Smart Technology, Beijing, China). The archwires consisted of 0.016- and 0.016 × 0.022-in stainless steel (Ortho Organizers, USA) or 0.016- and 0.016 × 0.022-in nickel titanium (Beijing Smart Technology), depending on the treatment phase. The sampling was performed at times immediately before initiation of treatment and 6 and 12 months later.

Saliva Collection

The patients were instructed, orally and in written, to avoid consumption of a given list of foods rich in nickel and chromium from 24 h prior to the next visit, which was scheduled in the morning. They were also told to avoid eating/drinking and brushing/mouth rinsing with fluoridated products from the night before the next visit [14].

The participants were asked to rinse their mouth for 30 s with distilled water, then wait for 2 min and eject the unstimulated saliva into nickel- and chromium-free sterile 2-ml polypropylene tubes (Vitalab, Germany). The specimens were stored in a low-temperature (−20°C) freezer by an operator wearing talc-free disposable vinyl gloves [13, 18]. Within maximum a month, the tubes were shipped to the Chemical Analysis Department of the Atomic Energy Organization for electrothermal atomic absorption spectrophotometry (at 0.01-μg/L accuracy limit) using a calibrated device (AA280Z GTA120, Varian, Mulgrave, Australia). After dilution with 65 % nitric acid, the content of each tube was examined twice. In the case of any inconsistency between the values of the two specimens from each tube, the values would be dismissed and the process would be repeated with two new specimens from that tube. These procedures were repeated at the sixth and 12th months of treatment.

Statistical Analysis

The sample size was determined as 3 × 20 specimens (for each metal) in a three-level repeated-measures design to obtain a test power >0.8 (effect size ≥0.3 μg/L, α ≤ 0.05). According to a D’Agostino–Pearson omnibus normality test, only the distribution of nickel values was normal. A two-way repeated-measures analysis of variance (ANOVA) and a Bonferroni post hoc test were used to assess the time-dependent changes in the concentrations of both metals together. Afterward, a repeated-measures ANOVA was used to analyze the nickel values. The data pertaining to chromium were analyzed using a Friedman and a Wilcoxon matched-pairs tests. The level of significance was set at 0.05 for the ANOVA, Bonferroni, and Friedman tests. It was adjusted to 0.016 for the Wilcoxon test using the Bonferroni method.

Results

There were high interindividual variations in metal amounts (Table 1); however, both average ion levels showed slight, comparable alterations during 1 year (Fig. 1; Table 2). According to the two-way repeated-measures ANOVA, the difference between the amounts of the two metals was significant (F = 44.81, P = 0.000) and the total change in concentrations of both ions was marginally significant (F = 2.59, P = 0.081). Nevertheless, the time × ion interaction was not significant (F = 0.034, P = 0.967), meaning that the effect of time was similar for leachability of these metals. The Bonferroni post hoc test showed that the differences between nickel and chromium were significant at all intervals (all three P values = 0.000).

Table 1 Nickel and chromium levels of the 20 patients at the three intervals
Full size table
Fig. 1
figure 1

Changes in mean (95 % CI) salivary ion levels in time (micrograms per liter)

Full size image
Table 2 Descriptive statistics of nickel and chromium levels at different intervals
Full size table

Mean salivary nickel had a tenuous increase after 6 months, which dropped to levels below control, after 1 year (Fig. 1; Tables 2 and 3). These fluctuations were not statistically significant according to the repeated-measures ANOVA (F = 0.774, P = 0.468). Nonetheless, the 95 % CIs pointed to some differences between the third interval with the other intervals (Table 3).

Table 3 Comparisons between the values observed at different intervals and the results of Wilcoxon test
Full size table

Mean chromium value followed a similar trend, increasing slightly at first, but dropping to sub-pretreatment concentrations after 12 months (Fig. 1; Tables 2 and 3). The alterations in chromium amounts over time reached the level of significance [P = 0.021 (Friedman)]. The decrease in the average chromium amount of the 12th month was statistically significant compared to the control (Table 3). The 95 % CIs implied that the difference between the third and second intervals might be as well generalizable (Table 3).

Discussion

No significant overall changes were seen among nickel values in this study. Slight but significant increases have been shown in other studies only immediately after appliance placement [9, 22], or 1 month later [14]. The large interindividual variations seen in this study, especially in nickel values, might be explained by the differences in saliva composition and pH influenced by various physiologic and environmental conditions such as the time of day, diet, health, and psychic conditions, in addition to adhesion of nickel to epithelial cells, bacteria, and salivary macromolecules [1, 9, 10, 13, 18]. Interindividual variations and mean values peaked at the sixth month of this study, possibly because the maximum release of metals is at the first 4 to 5 months [15]. Except for the high variation found in nickel concentrations, average nickel and chromium values exhibited similar trends, both falling somewhat (about 1.5 μg/L) below normal levels after 1 year of therapy. These fluctuations were small but became significant in the case of chromium. Higher variations in nickel values might preclude the difference from reaching the statistical significance threshold as 95 % CIs still indicated some changes over time.

Some of previous reports showed small decreases in nickel [9] or chromium levels [9, 14]. In most of other researches, either no significant alterations occurred in ion concentrations [13, 14, 16, 18, 22, 25] or the ions were considerably elevated in the treatment group compared with control [9, 10, 26]. However, in contrast to all other retrospective long-term or prospective short-term studies, a statistically significant decrease was observed in chromium values during the 12-month course of this study. The similarity of the experimental conditions at each interval, comparatively low variations observed at the third interval, and the fact that this reduction was similarly seen in the two ions implied this finding was unlikely an artifact. The authors suggest that such a small but generalizable decline might be attributable to increases in uptake of these ions by plaque microorganisms (or oral mucosa cells), which can accumulate metal ions at a significant rate by complexing with glycoproteins or previously absorbed ions existing in the biofilm [1, 3, 7, 8, 18]. Therefore, it might be enhanced by increases in the amount of such molecules or bacteria in the biofilm, which correlate to thickening of the plaque caused by difficulties in maintaining proper oral hygiene during orthodontic treatment [18, 19]. Simultaneous increases in oral ion absorption and decreases in metal ion release after the initial treatment stage [1, 8, 15] might partly justify the small but statistically significant decrease in salivary ion amounts below control values. However, no similar reports or designs were available to compare the results. Most of the available three retrospective long-term studies indicated the absence of significant changes in ion levels [10, 16, 18], just like what was seen in nickel values in this study, although an increase in nickel was observed in one research [10]. Nevertheless, with better controlling for the confounders in this study, one of the ions revealed statistically significant changes, eventhough still not clinically substantial.

Consistent with the findings of other studies [9, 10, 13, 14, 16, 18, 22, 25], the present study showed that the increase in nickel and chromium concentrations caused by orthodontic appliances was never comparable to dietary amounts (100–800 μg/day for nickel and 50–280 μg/day for chromium) [5, 13, 18]. Therefore, it seems that they might never reach toxic thresholds during orthodontic therapy. This might be, however, a false safety assurance because low but chronic release of metal ions as the result of corrosion can produce inflammation, alterations of cellular morphology/metabolism, and DNA damage/instability [3, 8, 15, 16]. On the other hand, it should be taken into account that a majority of such corrosion products are unlikely to be carcinogen or toxic [8, 27], and the DNA damage might be reversed after removal of orthodontic appliances [15]. Overall, the main concern with these ions in orthodontics is hypersensitivity, as nickel and chromium are, respectively, the first and second most common sources of type IV sensitivity worldwide [6, 19, 24, 26]. The allergic stomatitis caused by orthodontic therapy characterizes by several signs/symptoms including lip desquamation, gingival hyperplasia/inflammation, periodontitis, burning sensation in the mouth, and metallic taste [6, 19, 27]. Nickel can also cause damage to periodontal tissues in sensitive patients, which this phenomenon might be more prevalent than its sensitizing influence [19].

The baseline values observed in this study were within the ranges reported earlier, mostly being rather at low levels (0.64 [9]–3.9 μg/L [in the current study] for chromium and 0.53 [13]–11.9 [10] μg/L for nickel). However, two studies revealed much higher amounts for chromium (39 μg/L [26] and 61 μg/L [25]) and nickel (46.7 μg/L [26] and 55 μg/L [25]). This large difference is attributable to the variations in salivary composition, quality of food, bacterial colonization, galvanic currents produced [10, 13, 23, 24], and sampling methods (e.g., stimulated vs. unstimulated saliva collection) [16]. For instance, it was shown that rinsing with distilled water can noticeably diminish salivary metal ion concentration [16]. Therefore, dissimilar periods of waiting between rinsing and saliva collection (e.g., 2 min in this study and that of Fors et al. [18] vs. 5 min in some other studies [14, 26]) might be another factor corresponding to detecting different ion amounts. Moreover, stimulation of saliva secretion in some studies might contribute to higher ion amounts [13, 26]. In this study and some others, nickel level was found to be by far higher than chromium either in saliva or mucosa cells [3, 8–10, 14, 15]. Some investigators, however, showed greater concentrations for chromium before inserting the appliances, but not later [26] and in some other studies, chromium was always more elevated [13, 16, 25]. The difference between chromium and nickel might be due to diets and other sources of exposure, in addition to dissimilar effects of methodologies (such as rinsing) on the two ions. For example, Eliades et al. [16] showed that rinsing can affect nickel amounts more than chromium.

This and previous in vivo studies were limited by some factors. Many variables are involved upon measuring salivary metal release in vivo, which can undermine the accuracy of the results of clinical studies [3, 10, 13, 23, 24]. Moreover, the short sampling period in these studies and continuous flow of saliva might not give time sufficient for release of a notable dissolution of nickel/chromium from the fixed appliances [13]. Besides, it is not known whether ionic release follows a steady state pattern and that if the ion concentrations established with momentary sampling at a specific time could represent the full-term rates of ion discharge [3, 16]. It could not be true because most of these studies have measured ions in saliva samples collected at least 1 or 2 h after a meal or before breakfast (such as in the present study and some others [14]) in order to standardize the salivary flow and avoid influence of diet on evaluated ions. Nonetheless, the peak of metal ion release might be shortly after consuming food because of reduction in pH [18]. So the actual release rates of these ions might be closer to what has been found in controlled laboratory conditions which have demonstrated much higher rates than what is observed by current sampling approaches in vivo, although again far below dietary amounts [2, 13, 18]. Moreover, a clinical trial design with a separate control group could exert more control over the confounder variables. However, this was out of the time and budget projected for this 1-year prospective study with a rather long list of exclusion criteria, as none of the other prospective studies had assessed patients for more than 2 months. All earlier studies on salivary, systemic, or mucosal cell metal ion levels were limited to one or more of the following constraints, including adopting retrospective cohorts [8, 10], short follow-up durations [9, 18, 20, 22, 24, 27], reducing sample sizes [3, 9, 22, 24, 27], using longitudinal designs without control groups [9, 22, 24, 27], or adopting descriptive methodologies [14]. Finally, unlike almost all other studies [8–10, 14, 18, 20, 22, 24, 27], instead of only relying on statistical significance, confidence intervals were computed and interpreted as well to avoid the effect of sample size compounded with interindividual variations on the generalizability and comparability of the findings.

Conclusions

Within the limitations of this study, it was shown that the increase in each of the studied ion concentrations at the sixth month of treatment was insignificant. It was interesting to observe a statistically significant (although still small) decrease in mean chromium level at the 12th month of treatment. The effect of time was similar for leachability of these metals, and the mean nickel value as well decreased to a similar extent, though without the reduction becoming statistically significant. All changes found in this study were inconspicuous in terms of clinical significance, which might confirm the safety of long-term orthodontic treatment, as far as the ion amounts, not their dosage-independent effects, are of concern.